update

examples/vm_sound_classification/run.sh

@@ -12,8 +12,8 @@ sh run.sh --stage 2 --stop_stage 2 --system_version windows --file_folder_name f
 E:/Users/tianx/HuggingDatasets/vm_sound_classification/data/wav_finished/id-ID/wav_finished/*/*.wav" \
 --label_plan 4
 
-sh run.sh --stage 2 --stop_stage 2 --system_version centos --file_folder_name file_dir --final_model_name vm_sound_classification2-ch4 \
---filename_patterns "/data/tianxing/PycharmProjects/datasets/voicemail/*/wav_finished/*/*.wav" --label_plan 2
+sh run.sh --stage 0 --stop_stage 5 --system_version centos --file_folder_name file_dir --final_model_name vm_sound_classification8-ch32 \
+--filename_patterns "/data/tianxing/PycharmProjects/datasets/voicemail/*/wav_finished/*/*.wav" --label_plan 8
 
 "
 

toolbox/torchaudio/augment/spec_augment.py

@@ -1,46 +0,0 @@
-#!/usr/bin/python3
-# -*- coding: utf-8 -*-
-"""
-https://github.com/wenet-e2e/wenet/blob/main/wenet/dataset/processor.py
-"""
-import random
-from typing import List, Tuple
-
-import torch
-import torch.nn as nn
-from torch.distributions import uniform
-
-
-class SpecAugment(nn.Module):
-    def __init__(self,
-                 aug_volume_factor_range: Tuple[float, float] = (0.5, 2.0),
-                 ):
-        super().__init__()
-        self.aug_volume_factor_range = aug_volume_factor_range
-
-    @staticmethod
-    def augment_volume(spec: torch.Tensor, factor_range: Tuple[float, float] = (0.5, 2.0)):
-        factor = uniform.Uniform(*factor_range)
-        factor = factor.sample()
-        spec_ = spec.clone().detach()
-        spec_ *= factor
-        return spec_
-
-    def forward(self, spec: torch.Tensor) -> torch.Tensor:
-        spec = self.augment_volume(spec, self.aug_volume_factor_range)
-        return spec
-
-
-def main():
-    spec_augment = SpecAugment()
-
-    spec = torch.randn(size=(1, 10, 4))
-    print(spec)
-
-    spec_ = spec_augment.forward(spec)
-    print(spec_)
-    return
-
-
-if __name__ == '__main__':
-    main()